Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add chrisblattman/claudeblattman --skill adversarial-reviewgit clone --depth 1 https://github.com/chrisblattman/claudeblattmanWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/chrisblattman/claudeblattman/adversarial-review)<a href="https://agentmods.dev/skills/chrisblattman/claudeblattman/adversarial-review"><img src="https://agentmods.dev/badge/skills/chrisblattman/claudeblattman/adversarial-review/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/chrisblattman/claudeblattman/adversarial-review"><img src="https://agentmods.dev/badge/skills/chrisblattman/claudeblattman/adversarial-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00045 | $0.00440 |
| Opus 5.5 | $0.00018 | $0.00176 |
| Sonnet 5.5 | $0.00009 | $0.00088 |
| Haiku 4.5 | $0.00005 | $0.00044 |
Grade A, and why
adversarial-review scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Adversarial review
Review the supplied work against its original purpose, evidence, audience, and constraints. Read references/host.md and references/panel-execution.md and follow that complete execution contract and the shared Quality and revision contract in council-synthesis.md. The shared scripts/council.py infers this review workflow from its installed directory.
The public default is four fresh local reviewers: skeptic, pre-mortem, completeness checker, and Chief of Staff, followed by a separate fresh synthesis worker. The host and authoring workers do not count as fresh reviewers. Use the review guidance in each bundled persona. If saved or explicitly requested provider seats apply, use those exact seats instead and report the count clearly.
Reviewers independently identify material factual errors, unsupported claims, implementation defects, missing requirements, usability problems, and consequential risks. Give evidence or locations, explain consequences, and propose the smallest sufficient fix. Separate demonstrated defects from uncertain concerns and personal preferences. Preserve the user's intended voice; do not rewrite merely to impose a house style. If there is nothing material to fix, say so. This is not an instruction to invent objections.
The synthesizer weighs evidence rather than counting votes, retains important minority findings, and returns a prioritized set of changes and open decisions. A missing required reviewer makes the result incomplete even when useful synthesis is available. Never call a partial panel a complete review.
Return findings first. Edit the artifact only when the user has authorized revisions, and verify selected changes in the actual final artifact. Preserve accepted repairs, rejected suggestions and held questions in a short working summary. Do not launch another review, multi-round debate, or automatic council after revising. A further review requires a new request or an already commissioned stage and a new archive. If no existing artifact or proposal is available, resolve that missing input or route an explicit development request to Council once.
What ships with it
7 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 17 lines · 45 tokens per session scan A fb433543f8a1
adversarial-review is a skill published in the GitHub repository chrisblattman/claudeblattman (474 stars, last pushed today), licensed MIT. It adds 45 tokens to every session and 440 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-30.
Other skills, from other repositories
generic-react-code-reviewer
Review React/TypeScript code for bugs, security vulnerabilities, performance issues, accessibility gaps, and CLAUDE.md workflow compliance. Enforces TypeScript strict mode, GPU-accelerated animations, WCAG AA accessibility, bundle size limits, and surgical simplicity. Use when completing features, before commits, or…
generic-fullstack-code-reviewer
Review full-stack code for bugs, security vulnerabilities, performance issues, accessibility gaps, and CLAUDE.md compliance. Enforces TypeScript strict mode, input validation, GPU-accelerated animations, and design system consistency. Use when completing features, before commits, or reviewing pull requests.
generic-static-code-reviewer
Review static site code for bugs, security issues, performance problems, accessibility gaps, and CLAUDE.md compliance. Enforces pure HTML/CSS/JS standards, minimal page weight, mobile-first design. Use when completing features, before commits, or reviewing changes.
low-code-platforms
Low-code and internal tool platforms including Retool, Supabase, Appsmith, Tooljet, n8n, and Zapier. Use when building admin panels, internal tools, workflow automations, or integrating low-code platforms with custom code.
agent-desktop
Control native desktop apps with agent-desktop accessibility snapshots and actions on macOS, Windows, or Linux. Use for desktop UI tasks, never browser automation or web research.
ai-architect-review
Review the architecture of a system that calls a language model. Use when adding an agent loop, a retrieval path, a tool surface, or an MCP server to a codebase; when a system that worked in a demo is being prepared for production; or when asked to audit, review, or plan AI/agent architecture. Checks the four…